Jump to content

Arran

MTA Contributors
  • Posts

    583
  • Joined

  • Last visited

Member Title

  • MTA Contributor

Details

  • Gang
    CIT
  • Location
    UK

Recent Profile Visitors

3,991 profile views

Arran's Achievements

Mack

Mack (28/54)

25

Reputation

  1. I don't think that saying someone is talented is negative or defamatory. Anyway, on topic, from my experience with your work I'm sure you have the talent to complete this project and probably anything else like it.
  2. This sounds like a really big project and you seem to like major projects, doing loads of work on them, but then lose interest near the end. Which is a real shame that so much talented work goes to waste like MTA: Eir. There are much smaller issues that can have a big benefit to all players, like how client FPS is so low. I traced one cause of low client FPS to sounds: https://github.com/multitheftauto/mtasa-blue/issues/1067 Surely finding the root cause (see the newest comment posted on the issue as to how we might be able to do that) and then fixing it could be a fraction of the work of some big project like this, but would benefit every single of the tens of thousands of MTA players.
  3. After watching mtaqa Twitch stream hosted by @Woovie which included some time spent playing on CIT I realised that there are some obvious flaws which never get reported on our forum especially things that affect new players as our forum is unlikely to ever get feedback from brand new players. From watching Woovie's stream these problems were pointed out and have been fixed: Tutorial UI having a line that wasn't straight which was what he switched to full screen mode to try fix. Doubled the strength of the marker shader texture so it's more visible. Made the text when entering certain markers much more visible. So I thought that rather than using this topic to advertise, it would be better to receive feedback from anyone whose just tried the server out. This topic isn't for you to rant about some nonsense made up by dramatists.
  4. Currently to test a pull request: 1. 4 minutes to download (appveyor download today was only going 400 kb/s so takes almost 4 minutes to download) 2. 1 minute to copy all the files into the MTA installation. 3. 30 seconds to modify the meta.xml each time. 4. 1 minute to launch client and server and join the server. So at the moment it takes over 6 minutes just to get the PR build launched to test. Can't PR's all just get merged into a testing branch so can test all PR's in 1 go? This is especially important because if a PR is tested 1 by 1 we're unlikely to notice any rare bug in the PR that could take hours of playing random stuff to come up.
  5. " i got rights from Brian and arran " - Uhh no you haven't? Some scripts got leaked many years ago and many servers have been using them for years. MTA team are not responsible for dealing with failures in trust by server admins. Also they just decompiled some client side scripts which seems to have been done many years ago. And if a server has to use stolen scripts they will obviously never be a server worth playing on.
  6. Arran

    Open Source

    I joined discord. I've already tested the only PR I can see (fast_lookup_byindex) on nightlytest and posted the results of that already. I'm aware that I can download them on github at the checks => appveyor => details => artifacts. @qaisjp how about some tags on PR's like a "needs testing" tag so once an MTA developer has looked at the code and would be happy to merge it if it's been tested add the "needs testing" tag and then someone (probably me) can test it and then post back the test results and I guess then I could change the tag to "tested".
  7. Arran

    Open Source

    Anyone can contribute as you can see here: https://github.com/multitheftauto/mtasa-blue/pulls Unfotunately it seems that not enough people are given commit access to merge pull requests which is preventing all these pull requests getting merged. I counted 13 different people in those 42 pending pull requests who do not have access to merge pull requests who may want to be doing even more pull requests if they got merged. I made source patches in the past but I didn't feel like doing more until the existing ones had been looked at and by the time they did get looked at I'd lost interest / forgot what I'd learned. The same could apply to these people, they might want to do more but they don't see the point if even their already done stuff may never get merged. I'm not really sure what the problem is because correct me if I'm wrong but an MTA dev just needs to briefly look at the changed code to make sure it's not dodgy and then the community like me can spend the time doing the testing? Which I would do and have done. Or at least trust a few more people to make commits like the ones who have submit the most PR's. No because obviously you could put cheats in your version. You can only use a modified / a custom build version of MTA on modified / custom built versions of the server.
  8. MTA server has 4 threads: Main, sync, raknet and database. Although this doesn't mean it could completely use up 100% of a CPU's power, it doesn't need to because there are no servers popular enough to be affected by this. The record for online players on 1 server is 1400 and that server did not even lag during that insane peak, this was thanks to all the effort ccw made optimizing MTA server and the semi multi threaded support it has. Also efficient scripts help which again is thanks to ccw's performancebrowser which allows scripters to see what needs optimizing. It is client performance that is much more of a problem now. Many of us have modern PC's that can play GTA V on high and have stable 60 FPS yet on MTA servers with many objects, vehicles and players around we can get as low as 30 FPS. Those with older PC's don't stand a chance.
  9. I can't remember so I tested again, I copied my MTA installation, put the build files to over write, had to give users edit access to folder, server and client worked fine. Used the custom animation build this time.
  10. Sounds like you need a better VPS provider. OVH GAME dedicated servers are almost completely immune to DDOS but I guess if you have a VPS that's out of your budget.
  11. Ahh, I've tried one and my client crashes when joining the local server, gonna try another, 79 MB from appveyor.com takes 5 minutes. (Was trying "Add setElementCollidableWithType #181") Also I just realised there isn't actually a 1.6 section on nightly.mtasa.com so does that mean master is 1.5.5? As someone whose been testing next release version since 1.0 it doesn't seem right that there is no 1.6 because... well all the major additions that MTA has had in the past, needed it so we could test them and for when there's non backward compatible changes. Second attempt I'm now getting "Multi Theft Auto has not been installed properly, please reinstall." (Was trying "Added dynamic ped ID allocating(based on lopezloo pull request) #151")
  12. You're talking about https://github.com/multitheftauto/mtasa-blue/pulls right? If so then I guess that is for developers only because I don't see what you describe.
  13. I've been checking MTA commits multiple times a day for about 8 years, you know I updated wiki and mantis regarding those merged pull requests? If you're planning on merging more that's great, just tell me what I can do to help, such as if there was a way for me to test the pull requests.
  14. Some pull requests are over a year old, there's 23 pending, and a lot of them have this tag: "pr:needs-testing" I would be happy to test these if there was a way to test them without having to go to all the effort of compiling my own build. Correct me if I'm wrong but 'master' is 1.6 and therefore basically all pull requests could be merged to master right now (or make a branch where all pull requests are merged to) and then we'd be able to test it all with the 1.6 nightly and then the whole community would have the ability to test all pull requests, then we can confirm whether or not the pull requests work. 1.6 is a long way off so there'd be plenty of time for any bugs to get discovered and if we're given a few weeks notice before a 1.6 release we can do general testing to see if any random bugs have popped up. A lot more pull requests might actually get made if they got merged, back when I made some source pathces, it was demotivating to see my patches get ignored, then I just lost interest. The people with pending pull requests from months ago are probably feeling the same thing. To be honest, all the pull requests combined could actually be good enough for a new release especially when there's "Added dynamic ped ID allocating" and if there was dynamic object ID allocating I'd be begging for 1.6 to get released. All you guys gotta do is check their code to make sure it looks legit, then I can do all the ingame testing.
×
×
  • Create New...